Download as pdf or txt
Download as pdf or txt
You are on page 1of 17

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/285743185

Amur tiger: a case study of living on the edge

Article · January 2010

CITATIONS READS

55 2,333

10 authors, including:

Dale Miquelle John Goodrich


Wildlife Conservation Society Panthera
179 PUBLICATIONS   4,214 CITATIONS    91 PUBLICATIONS   2,508 CITATIONS   

SEE PROFILE SEE PROFILE

Philip Andrew Stephens Linda L. Kerley


Durham University Zoological Society of London, UK
121 PUBLICATIONS   6,378 CITATIONS    28 PUBLICATIONS   1,006 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Methods For Monitoring Tiger And Prey Populations View project

Tiger-Leopard Observation Network in China (TLON) View project

All content following this page was uploaded by Philip Andrew Stephens on 20 April 2016.

The user has requested enhancement of the downloaded file.


McDonald 13-McDonal-chap13 Page Proof page 325 28.8.2009 9:14pm

CHAPTER 13

Amur tiger: a case study of living on the edge


Dale G. Miquelle, John M. Goodrich, Evgeny N. Smirnov, Phillip A.
Stephens, Olga Yu Zaumyslova, Guillame Chapron, Linda Kerley,
Andre A. Murzin, Maurice G. Hornocker, and Howard B. Quigley

# Dale Miquelle AQ1

edge of a species’ distribution provides a lens


Introduction through which to understand the ecological con-
Although the stereotypic image of tigers usually straints imposed not only on that population but
places them in a steamy Indian jungle, in fact the the species as a whole. Because Amur tigers are al-
species occupies diverse habitats across a wide range ready stressed by extreme environmental conditions,
of latitudes. The Siberian or Amur tiger (Panthera we should expect them to be more sensitive to
tigris altaica) is the northernmost, and probably human disturbances. Thus, understanding the im-
faces the most severe suite of environmental con- pact of humans on tigers at the northern fringe of
straints. Although tigers in general have moderate their range provides a basis for conserving not only
to low levels of genetic diversity (Wentzel et al. Amur tigers, but all subspecies.
1999; Luo et al. 2004), and the Amur tiger appears We hypothesize that the primary constraints on
to be particularly homogenous (Luo et al. 2004; Rus- tiger distribution and abundance are (1) prey distri-
sello et al. 2004), its ability to adapt to the hostile bution and abundance and (2) mortality and habitat
environments of north-east Asia demonstrate the loss caused by people, factors considered critical for
behavioural and physiological plasticity of this spe- other large felids as well (e.g. Woodroffe [2000], Car-
cies. Understanding how a population survives at the bone and Gittleman [2002], Dinerstein et al. [2007],
McDonald 13-McDonal-chap13 Page Proof page 326 28.8.2009 9:14pm

326 Biology and Conversation of Wild Felids

and Hayward et al. [2007]). In this chapter we inves-


Description of Amur tiger range
tigate how the abundance and distribution of prey
influences density, home-range size, reproductive The present geographic range of Amur tigers is
parameters, and survival rates of Amur tigers. By primarily confined to the Russian Far East pro-
comparing characteristics of Amur tigers with those vinces (Krais) of Primorye and Khabarovsk (Fig.
of more southern populations, especially Bengal 13.1). There are probably less than 400 adult and
tigers (Pa. t. tigris), we reveal how varying ecological subadult tigers in Russia (Miquelle et al. 2007),
constraints influence conservation strategies. with about 95% occurring in the foothills and

Figure 13.1 (a) Sikhote-Alin and east Manchurian (Changbaishan) mountain ecosystems, and (b) current distribution of
tigers in north-east Asia, based on preferred habitat and track locations from winter surveys in 2005 in the Russian Far East
(Miquelle et al. 2007), and in Northeast China, based on verified reports and surveys in 1998 and 1999 (Jiang 2005; Yu 2005).
McDonald 13-McDonal-chap13 Page Proof page 327 28.8.2009 9:14pm

Amur tiger: a case study of living on the edge 327

main ranges of the Sikhote-Alin Mountains. There In both Sikhote-Alin and East Manchurian forest
is also an isolated population of 20 or so tigers ecosystems, peak altitudes are generally 500–800 m
distributed within the East Manchurian (or Chang- above sea level, few reaching 1000 m or more. Col-
baishan) mountain system of eastern Jilin and lectively, these two Tiger Conservation Landscapes
southeastern Heilongjiang Provinces of China and (TCLs; Sanderson et al. 2006) represent a zone where
Southwest Primorski Krai, Russia. Today, most East Asian coniferous–deciduous forests and the bo-
tigers in the East Manchurian Mountains occur real forests merge, resulting in a mosaic of forest
on the Russian side, but the majority of available types that vary with latitude, elevation, topography,
habitat is on the Chinese side. Historically, the proximity to the Sea of Japan, and past history.
largest tracts and most productive tiger habitat Over 70% of Primorye and southern Khabarovsk is
(and therefore the largest numbers of Amur tigers), forest covered and, despite more extensive clearing,
no doubt were found in Northeast China. When large tracts of forest also remain in Northeast China
tigers nearly disappeared from the Russian Far East (Fig. 13.2). The original dominant forest was com-
in the 1940s (Kaplanov 1948; Miquelle et al., in posed of a mixture of Korean pine (Pinus koraiensis)
press), dispersal from China probably contributed and deciduous tree species including birch (Betula
to their recovery in Russia (Heptner and Sludskii spp.), basswood (Tilia spp.), aspen (Populus spp.) and
1992). However, the development of Northeast in the north and at higher elevations, conifers such as
China, starting in the late nineteenth century and spruce (Picea spp.), fir (Abies spp.), and larch (Larix
increasing dramatically in the 1940s, led to habitat spp.). Most of these forests have been selectively
loss and intensive hunting (Ma 2005) so that by logged in the past, and human activities, in associa-
the 1990s the few tigers still reported in Northeast tion with fire, have resulted in conversion of many low
China were probably mostly dispersers from Russia elevation forests to secondary oak (Quercus mongolica)
(Yu 2005). and birch (Be. costata, Be. lanata, and others) forests. AQ3

Pine and deciduous forests

Conifer and alpiine/tundra


Open meadows, agricultural
fields, and settlements
Northern boundary of pine
and deciduous forests
Historic northern boundary
of tiger distribution
Russian-Chinese boundary

Figure 13.2 Northernmost distribution of tigers, reconstructed from Heptner and Sludski (1992), and distribution of the
mixed-pine (Pinus koraiensis and Pinus sylvestris) broad-leaved and pure deciduous forests, based on data from two sources:
a vegetation map of the Amur Basin developed under V.H. Sochavi (Soviet Academy of Sciences, 1968) and a map of habitat
types of Sikhote-Alin by V.V. Ermoshin, A.A. Murzin, and V.V. Aramilev (Institute of Geography, Far Eastern Branch of the
Russian Academy of Sciences) interpreted from Landsat 5 satellite imagery.
McDonald 13-McDonal-chap13 Page Proof page 328 28.8.2009 9:14pm

328 Biology and Conversation of Wild Felids

We have intensively studied Amur tigers (from Sikhote-Alin and much of the eastern Manchurian
1992 to present) in the Sikhote-Alin Zapovednik (Re- Mountains, but are rare in higher altitude spruce–fir
serve), an IUCN Category I protected area which forests. Red deer become rarer in the southern por-
straddles the Sikhote-Alin Mountains (Fig. 13.1) tions of this region, where they are replaced by sika
and is the largest protected area containing tigers in deer (C. nippon) which reach their northern limits in
north-east Asia. central (inland) and northern (along the coast) Pri-
The climate in this region is monsoonal, with 75– morye. Roe deer are often associated with forest
85% precipitation (650–800 mm in central Sikhote- ‘edge’ habitats, but also occur in forests throughout
Alin) occurring between April and November. The the region. Manchurian moose (Alces alces came-
January monthly mean temperature is 22.6 C on loides) (281 kg; Bromley and Kucherenko 1983)
the inland side of the central Sikhote-Alin Moun- reach the southern limits of their distribution in
tains, but the Sea of Japan moderates temperatures central Sikhote-Alin Mountains, and are sparsely
(and snow depths) on the coastal side of the range, distributed in the inland boreal forests, but more
when the January monthly mean is 12.4 C. The abundant further north. Musk deer (Moschus moschi-
frost-free period (which influences plant productivi- ferus) are most commonly associated with conifer
ty) varies between 105 and 120 days/year. Snow forests, but can be found in a wide variety of forest
depth varies greatly by locale and year, but averages habitats. Goral (Nemorhaedus caudatus; 32 kg; Vo-
22.6  2.9 cm in mid-February (60-year mean) in the loshina and Myslenkov 1994), a rare wild goat, are
inland central Sikhote-Alin and only 13.7  3.5 cm mostly limited to coastal cliff habitat. Unique to
on the coast (data from Primorye Weather Bureau). north-east Asia is the absence of wild bovids in the
However, heavy snowfalls, most common in March, spectrum of available prey for tigers (Table 13.1).
can have dramatic impacts on ungulate and tiger Although bovids do not always represent a large per-
populations. centage of the items selected, because of their size
they can provide a significant proportion of the total
biomass consumed by tigers (Karanth 1993).
Representing both potential competitors and prey,
Food habits and prey preferences brown bear (Ursus arctos) and Asiatic black bear
(U. thibetanus) are common in north-east Asia.
of Amur tigers
Given a relatively broad range of potential prey, it
By necessity tigers have developed the capacity to is not surprising that Amur tigers have been reported
prey on a wide range of species. However, there is eating everything from eagles to seals to brown
doubtless a suite of species and body weight ranges bears. However, of 720 kills made by tigers found in
that tigers prefer, which likely maximize the differ- and around Sikhote-Alin Zapovednik between 1962
ence between energy expended and gained in acquir- and 2003, 94% were ungulates, with red deer the
ing prey, tempered by the risk associated with primary prey item (54%) followed by wild boar
capture (Hayward and Kerley 2005). Tigers are likely (28%; Miquelle et al. 2005c). Roe deer accounted for
to forage optimally when taking the largest prey that only 6% of the diet. Sika deer and musk deer ac-
can safely be killed (Pyke et al. 1977; Hayward and counted for 3% and 1% of the kills, respectively,
Kerley 2005), often ungulates their own size or larger while bears (both Himalayan and brown bears)
(Sunquist and Sunquist 1989). The list of primary accounted for <3% of the kills.
prey across Asia usually includes a suite of large-and Data from six other sites across tiger range in Rus-
medium-sized cervids, as well as suids and bovids sia confirm that red deer and wild boar are the two
(Table 13.1). key prey species (63–92% of kills, collectively) and
The ungulate complex of north-east Asia is repre- that combined with the two medium-sized cervids
sented by seven species, with red deer (Cervus ela- (sika and roe deer), these four ungulates comprise
phus; for body weights see Table 13.2), wild boar 81–94% of kills (Miquelle et al. 1996). Kill data some-
(Sus scrofa), and Siberian roe deer (Capreolus pygargus) times under-represent smaller prey items but ana-
the most common: all are found throughout the lyses of scats in Sikhote-Alin Zapovednik also
Table 13.1 Summary of ungulate assemblages in tiger range.

Huai Kha Taman Sikhote-Alin


Chitwan, Kanha, Nagarahole, Sunderbans, Khaeng, Negara, Zapovednik,
Type Nepala Indiab Indiac Bangladeshd Thailande Malaysiaf Javag Sumatrah Russiai

Large Present Present Present Formerly Present Present Present Present Present
deerj
Medium- Present Present Present Present Formerly Absent Absent Absent Present
sized
deerk
Small Present Present Present Present Present Present Present Present Present
deerl
Wild pigsm Present Present Present Present Present Present Present Present Present
Large Present Present Present Formerly Present Present Present Present Absent
bovidsn
a
Sunquist (1981); b Schaller (1967); c Karanth (1993); d Heindrichs (1975); e Seidensticker (1986); f Kawanishi (2002); g Rabinowitz (1989); h O’Brien et al. (2003) and Seidensticker and
McDonald 13-McDonal-chap13 Page Proof page 329

Suyono (1980); i this book.


j
Large deer-Al. alces, Ce. elaphus, Ce. unicolor, Ce. duvauceli, Ce. schomburgki, and Ce. timorensis
k
Medium-sized deer-Ce. nippon, Ca. capreolus, Axis axis, Ax. porcinus, and Ce. eldii
l
Small deer-Muntiacus spp., Moschus moschiferus, and Tragulus spp.
m
Wild pigs-S. scrofa, S. verrucosus, and S. barbatus
n
28.8.2009 9:14pm

Large bovids-Bos gaurus, Bos frontalis, Bos javanicus, Babulus babulis


McDonald 13-McDonal-chap13 Page Proof page 330 28.8.2009 9:14pm

330 Biology and Conversation of Wild Felids

Table 13.2 Jacobs’ index as a comparison of preference by tigers for five ungulate species, based on
proportion of kills and proportion of available ungulate prey numbers (from average prey density) for nine 3-year
periods between 1966 and 2002 in Sikhote-Alin Zapovednik, Russia.

Jacobs 95% Proportion of prey Proportion of kills Adult female body


Species index CI numbers (availability) by tigers mass (kg)

Wild 0.749 0.094 0.054 0.292 86.0a


boar
Red 0.263 0.139 0.453 0.589 149b
deer
Sika 0.571 0.295 0.035 0.024 74b
deer
Roe 0.737 0.104 0.341 0.070 35b
deer
Musk 0.862 0.117 0.117 0.013 12c
deer
AQ4 Notes: Jacobs’ index (Jacobs 1974) is estimated as D = (r  p)/(r + p  2rp), where r is the proportion of total kills made up by a species and
p is the proportional availability of the prey species. D ranges from +1 to 1, where +1 indicates maximum preference and 1 maximum
avoidance.
a
Bromley and Kucherenko (1983)
b
Dalnikin (1999)
c
Prikhodko (2003)

confirmed that these four ungulates were the domi- west of the divide (Stephens et al. 2006). Because of
nant food items (86% of diet items in scat, compared small sample sizes in some years, we estimated pro-
to 90% of kills), although other prey, including both portions of each species killed during 3-year inter-
species of bears and smaller prey items such as bad- vals, and calculated the mean prey abundance for
gers (Meles leucurus), appeared more commonly in these same 3-year intervals. Because surveys are con-
summer scats. ducted in winter when some prey species are not
Given their relative size and abundance, as well as active (e.g. bears and badgers), we focused on the
the risks associated with capturing them, we would five most common prey species: red deer, wild boar,
predict that red deer and wild boar should be not roe deer, musk deer, and sika deer (moose are virtual-
only the most common, but also the most preferred ly absent within this study area). Jacobs’ index was
prey items for Amur tigers. To test this, we used kill used as an index of preference (Hayward and Kerley
data from Sikhote-Alin Zapovednik for the years 2005).
1962–72 (from Gromov and Matyushkin [1974]) The results (Table 13.2) reveal that red deer and
and 1980–2003 (from Miquelle et al. [1996, 2005c]) wild boar, the two largest ungulates, are the most
to estimate percentage of occurrence of various prey preferred with the smaller cervids—musk, roe, and
in the diet, and used yearly surveys of ungulates sika deer—clearly not preferred. However, given that
based on winter track counts (1966–2003) converted wild boar are smaller than red deer (Table 13.2), it
to numbers (Stephens et al. 2006) to provide an esti- would be expected, based on body mass alone, that
mate of proportional abundance. We focused on that they would be less preferred, when in fact, the oppo-
portion of the Zapovednik east of the Sikhote-Alin site is true (Table 13.2). Although wild boar, with
Divide to estimate prey densities, because that is their sharp tusks, are probably more dangerous to
where nearly all tiger kills were reported and because hunt, they may be easier to stalk than red deer (Yu-
prey densities and proportions vary greatly east and dakov and Nikolaev 1977). Groups of wild boar, with
McDonald 13-McDonal-chap13 Page Proof page 331 28.8.2009 9:14pm

Amur tiger: a case study of living on the edge 331

heads down and while foraging noisily in the leaf play a role in defining the northern limits of tiger
litter, are much easier to approach than red deer, and distribution, the coincidence of prey distribution,
are likely not as fast when attempting escape. This habitat types, and tiger distribution suggests clear
greater vulnerability may explain their preferred sta- linkages. Thus, although individual dispersing tigers
tus. Roe deer, though common, may be too small to have been reported more than 1000 km north and
be worth hunting. Sika deer numbers have expanded west of their historical range (Heptner and Sludski
dramatically in the coastal area of Sikhote-Alin Zapo- 1992; Kirilyuk and Puzansky 2000), stable popula-
vednik, and have largely replaced red deer as the tions are restricted to more southern regions where
most common cervid in the southern Sikhote-Alin more productive forest systems support prey densi-
Mountains. However, this trend, which may be ties high enough to allow for successful reproduction
related to global climate changes (Zaumyslova to occur (see below).
2000) may not be beneficial for tigers. Although The clear linkages between habitat types, prey dis-
sika deer appear capable of reaching higher densities tributions, and tiger distribution provide insight into
than red deer or wild boar (Gaponov 1991; Gaponov where tigers historically occurred in nearby China
et al. 2006) they are half the size of red deer, and so (Fig. 13.2) which coincides well with existing infor-
tiger-foraging efficiency may decrease where they are mation on their former distribution there (Ma 2005;
a primary prey. Despite this concern, tiger popula- Yu 2005). This information, in conjunction with
tions in southern Sikhote-Alin appear to be surviving recent data on where suitable habitat still remains,
and reproducing despite the changing suite of avail- provides a mechanism for defining potential tiger
able prey (Miquelle et al. 2007). recovery zones in Northeast China (Miquelle and
Zhang 2005).

What limits tiger distribution


to the north? Tiger density and home-range size
in relationship to prey density
The historical northern limit of permanent popula-
tions of tigers in Russia (from Heptner and Sludski In the absence of human disturbance, available prey
[1992]) coincides strikingly (despite imprecise biomass is expected to be the primary determinant of
knowledge of the tiger’s historical range) with the carnivore density (Carbone and Gittleman 2002;
northern limits of the pine-mixed deciduous forest Karanth et al. 2004c), a prediction supported by
complex of the Russian Far East (Fig. 13.2). This many field studies (van Orsdol et al. 1985; Stander
congruence suggests that these forests represent a et al. 1997c; Eberhardt and Peterson 1999; Fuller and
proxy for the northern limits of the tigers’ primary Sievert 2001; Hayward et al. 2007).
prey: wild boar and red deer (Fig. 13.3a and b). Al- In Sikhote-Alin Zapovednik, estimates of ungulate
though clearly neither species is wholly dependent and tiger numbers have been collected annually
on these forest complexes (both extend to complete- since 1966 (Smirnov and Miquelle 1999; Stephens
ly different forest ecosystems in Europe), their north- et al. 2006). Using average female body mass as a
ernmost limits in the Russian Far East extend only measure of body size in ungulate populations (with
slightly beyond these mixed pine and deciduous for- males weighing more, and subadults and young
ests. Indeed, the zones of highest densities occur, weighing less), we converted estimates of numbers
with a few exceptions for red deer, in the southern- of ungulates into total ungulate biomass for the Za-
most regions of the Russian Far East (Fig. 13.3a and povednik (excluding moose and goral, which repre-
b), wholly within the limits of this forest complex. sent a minor part of the total biomass, and only 1.4%
Although red deer (and to a lesser extent wild boar) of the tigers’ diet). A regression analysis revealed a
occur beyond those limits, their densities are likely strong linear relationship between the total available
too low to sustain a tiger population. Although other prey biomass and tiger density (Fig. 13.4). Because
factors such as extreme cold and deep snows likely the most preferred prey—wild boar and red deer—are
McDonald 13-McDonal-chap13 Page Proof page 332 28.8.2009 9:14pm

(a)
Red deer distribution in
Russian Far East
Lower densities
Higher densities
Northern boundry of Korean
pine-mixed deciduous forests
Historical northern boundary
of tiger distribution

China

n
0 500 Kilometers

a
p
a
J
N

f
o
a
e
S
DPR Korea
100 200 300 400 Kilometers

(b)
Wild boar distribution in
Russian Far East
Lower densities
Higher densities
Northern boundary of Korean
pine-mixed deciduous forests
Historical northern boundary
of tiger distribution

China
n

0 500 Kilometers
a
p
a
J

N
f
o
a
e
S

DPR Korea
100 200 300 400 Kilometers

Figure 13.3 Distribution of (a) red deer and (b) wild boar in the Russian Far East (reconstructed from Bromley and
Kucherenko [1983]) in relation to northernmost distribution of tigers and northern limits of the pine-mixed deciduous forest
complex (from Fig. 13.2).
McDonald 13-McDonal-chap13 Page Proof page 333 28.8.2009 9:14pm

Amur tiger: a case study of living on the edge 333

0.7 (a)
0.6 18
Tigers/100 km2

16

Tiger density (per /100 km2)


0.5
14
0.4
12
0.3
10
0.2
8
0.1 6
0 4
0 100 200 300 400 500 600
Prey biomass (kg/km2) 2
0
Figure 13.4 The relationship between prey biomass 0 1000 2000 3000 4000 5000 6000 7000 8000 9000
(kg/km2) and tiger population density (adult tigers/100 Prey biomass (kg/Km2)
km2) in Sikhote-Alin Zapovednik, Russian Far East (Y =
0.0009X + 0.623, r2 = 0.63, and p < 0.0001), based on (b)
data collected from 1966 through 2002. Ungulate 18
estimates are based on winter track counts converted to 16
animal abundance using the Formozov formula (Mirutenko 14

Tigers/100 km2
1986; Stephens et al. 2006), and tiger numbers are 12
based on an expert assessment based on interpretation 10
of relative track size, track age, and distance between 8
tracks (Smirnov and Miquelle 1999; Miquelle et al. 2006). 6
y = 0.003x + 4.2688
4
r 2 = 0.3703
2
also the most abundant, this relationship is similar 0
whether we include all ungulate biomass, or only 0 500 1000 1500 2000 2500 3000 3500
that of the preferred species. Preferred prey biomass (kg/km2)

However, when we compared a greater range of Figure 13.5 The numerical response of tigers (animals/
both tiger and prey densities from 13 sites across 100 km2) to changes in (a) total prey biomass (Y =
Asia (Miquelle et al. 2005c), a curvilinear relation- 13.683e1229.52/X) and r2 = 0.53) and (b) preferred prey
ship was revealed (Fig. 13.5). This Type II functional biomass (y = 0.003x + 4.2688, r2 = 0.37, and p = 0.08).
response has not been reported for any other large (Data from Schaller 1967; Tamang 1982; Thapar 1986;
Karanth 1991; Stoen and Wegge 1996; Smirnov and
carnivores, but accords with Karanth et al.’s unre-
Miquelle 1999; Karanth and Nichols 2000; Stephens et al.
stricted model (2004c) of tiger density and prey
2006.) The curvilinear relationship in (a) is represented by a
numbers (the fact that they used prey numbers, Michaelis–Menton function of the form Y = aeb/X, where Y is
rather than biomass, does confound the compari- tiger density, X is prey density, and a and b are constants.
son, but it seems unlikely that conversion to bio-
mass would dramatically change the relationship).
There are several possible explanations for why of species between 86 and 550 kg, with the lower
this relationship is not linear. The sample size was limit representing the weight of female wild boar (a
small (n ¼ 13), and methods and accuracy varied. preferred prey of Amur tigers), and the upper limit
There may also be errors in estimating prey density, representing the approximate weight of adult female
tiger density, or both. However, the parallel with gaur. The relationship based solely on prey within
Karanth et al.’s finding (2004c), which relied upon what is presumed to be the preferred weight range is
independent data and a more rigorous design, sug- linear (Fig. 13.5b), similar to other large carnivores
gests that measurement errors are not the explana- (Hayward et al. 2007), but is not strong and margin-
tion. Alternatively, the curvilinear fit might indicate ally non-significant (r2 ¼ 0.37 and p ¼ 0.08). Wheth-
over-representation of non-preferred species. To as- er this relatively weak linear relationship (Fig. 13.5b)
sess this, for 9 of the 13 sites for which it was possi- is a result of insufficient data is as yet unclear, but
ble, we recalculated available biomass for the range suggests that a curvilinear relationship (Fig. 13.5a)
McDonald 13-McDonal-chap13 Page Proof page 334 28.8.2009 9:14pm

334 Biology and Conversation of Wild Felids

600

Tigress home range size (km2)


500 Y = 36,605X 0.8776
r 2 = 0.987
400

300

200

100

0
0 1000 2000 3000 4000 5000 6000 7000 8000
Prey biomass (kg/km2)
Figure 13.6 A hypothetical relationship between prey biomass and home-range size of adult resident female tigers,
Y =aX(b/X), where Y is female home-range size, X is prey biomass, a = 18.41, and b = 205.506. This relationship assumes
that all tiger populations contain territorial females, and there are on average 2.5 individuals per female home range,
including an adult resident female, a third of a male (assuming males cover home ranges of 3 females), 0.7 cubs (e.g.
Amur tigers on average produce 1.4 cubs/year, but only half survive to 12 months), and 0.5 transients per home range.

may be more appropriate for tigers, providing a Goodrich et al. 2005) but slightly overestimating
tantalizing indication that other factors (perhaps so- home-range size for Chitwan National Park, Nepal
cial regulation, see below) may be influencing tiger (41 km2 vs. the actual estimate of 20.7 km2; Smith
densities when prey biomass is extremely high. et al. 1987). Across the range of tigers in Asia, female
Wherever they have been studied, tigers have de- home-range size appears to calibrate so as to provide
monstrated a spacing system in which females de- a similar prey biomass per home range. Thus, al-
fend territories that overlap little with neighbouring though home-range size of resident female Amur
females, and males defend territories that include tigers in Sikhote-Alin Zapovednik is, on average, 22
one to nine tigresses (Smith et al. 1987; Goodrich times larger than that of Bengal tigers in Chitwan,
et al. 2005). Females are expected to maintain home total prey biomass per female home range is approx-
ranges sufficiently large to ensure adequate resources imately equal (actually 1.2 times greater in Sikhote-
for rearing young, but as small as possible to avoid Alin than Chitwan), using data from Tamang (1982)
unnecessary defence costs (Smith et al. 1987; Sandell and Stephens et al. (2006) to calculate prey biomass.
1989). Consequently, female home-range size is pre- These results suggest that prey density is a driving
dicted to be inversely related to prey density (Sandell force in determining home-range size in female
1989), but there are too few data to test this for tigers. tigers, as predicted for solitary carnivores (Sandell
However, assuming there are 2.5 animals/female 1989), and consequently is of major importance in
home range in the average tiger population (assum- considering area requirements for viable populations
ing, in addition to the adult female, 0.3 adult males– of tigers. The curvilinear relationship in Fig. 13.6 also
female home range, 0.7 cubs/year/home range, and suggests that as prey biomass becomes very large,
0.5 transients), it is possible to derive a prediction for there may be a minimum home-range size for adult
female tiger territory size using estimates of prey females (10–20 km2), below which social regulation
biomass and tiger densities from Fig. 13.5a. This may start to influence tiger densities.
model (Fig. 13.6) provides reasonable estimates At the other end of the spectrum, Figs 13.4 and
where measurements of home range exist, correctly 13.5a suggest that tigers in Sikhote-Alin Zapovednik
predicting 445 km2 for tigresses in Sikhote-Alin Za- are close to the lower limits of prey densities at which
povednik Russia (vs. the actual estimate of 440 km2; tigers can exist. Without the constraints of rearing
McDonald 13-McDonal-chap13 Page Proof page 335 28.8.2009 9:14pm

Amur tiger: a case study of living on the edge 335

cubs, home ranges can continue to increase to en- 0.5


sure sufficient prey can be found, as demonstrated

Percentage of litters
by male Amur tigers which maintain home ranges 0.2

on average in excess of 1000 km2 (Goodrich et al.


0.5
2005). We suggest that, at 440 km2, female Amur
tigers within our study area are nearing the limits of 0.1
home-range size in which reproduction can suc-
cessfully occur. While prey densities decline further 0.05
to the north where tigers still occur (and presum-
ably home ranges increase in size), the logistics of 0

b
ar

r
ay

l
g

p
ct

c
Ju
Ap
capturing ample prey in the vicinity of immobile

No

De
Ja
Fe

Ju

Au

Se

O
M

M
young cubs, finding sufficient food to feed subadult
Figure 13.7 The percentage of 27 Amur tiger litters born
cubs still dependent on their mother (amounting in each month to 13 tigresses in and around Sikhote-Alin
to 2–4 times her individual energetic needs), and Zapovednik, 1992–2004 (Kerley et al. 2003, 2005).
the energetic demands of covering such large areas
may make significantly larger home ranges non-
viable for breeding females. demanding period for females would be when cubs are
15–18 months old. Male cubs at this time are already
larger than their mothers, but seem to be largely de-
pendent on them for food. For females giving birth in
Reproduction, survival rates,
July or August, the peak demands of feeding a family
and population persistence would occur between October and February of the
All northern ungulates demonstrate a birthing pulse next year. As demonstrated above (Table 13.2), tigers
in late spring and early summer that ensures that the prefer larger prey, but this preference is likely to be
most nutritious forage will be available to females to even stronger for females with large cubs. Small pack-
compensate for the energetic demands of lactation, ets of prey (e.g. newborn ungulates) would not provide
and provides the longest period of quality forage for sufficient food to feed a family, even if kills were made
growth of neonates before facing the rigors of a north- daily: we have observed that females with two to three
ern winter. Surprisingly, Amur tigers do not demon- yearling cubs can consume adult red deer in 24–36 h.
strate a similar birthing pulse in spring that might take By late autumn and early winter, calves of the year are
advantage of vulnerable young ungulates for feeding sufficiently large to provide reasonable volumes of
young cubs. In fact, while we have observed tigers meat, and are still relatively easy to catch. Also, hunt-
giving birth in all but 2 months of the year, the peak ing ungulates is probably easier for felids in winter,
birthing season is July and August (Fig. 13.7). Cougars when fresh tracks in snow provide clues to finding
(Puma concolor) in northern environments demon- prey (Yudakov and Nikolaev 1987), deep snow can
strate a very similar pattern (Cougar Management inhibit escape by prey, and soft snow acts as a muffler
Guidelines Working Group 2005). The period of great- for silent stalking. Thus, the shift in birthing season
est energy demands for female felids raising cubs is from spring and early summer, characteristic of Bengal
likely not when caring for newborns (even though tigers (Smith and McDougal 1991; Smith 1993) and
there is a high energy cost of lactation over the first captive Amur tigers (Seal et al. 1987), to late summer
few months), but when females must provide food for may be a strategy for females to ensure maximum prey
nearly full-grown subadults, which still rely on their biomass is available for their offspring prior to inde-
mother, and require as much or more energy (in the pendence and dispersal, when energetic demands of
case of male cubs) than she does. Our data suggest that the family unit are greatest.
Amur tiger cubs disperse from their natal home ranges Given their hostile environment, it might be ex-
on average at 19 months, and that they are moving pected that reproduction and survival rates of Amur
largely independently of their mother by 18 months tigers would be lower than those in more southerly
of age (Kerley et al. 2003). This suggests that the most populations. However, the scanty available data
McDonald 13-McDonal-chap13 Page Proof page 336 28.8.2009 9:14pm

336 Biology and Conversation of Wild Felids

Table 13.3 Comparison of reproduction parameters of Amur tigers in Sikhote-Alin Zapovednik, Russia (Kerley
et al. 2003), and Bengal tigers in Chitawan National Park, India (Smith and McDougal 1991; Smith 1993)

Bengal tigers Amur tigers


Parameter
n Mean 95% CI n Mean 95% CI

Age of first reproduction 5 3.5 — 6 3.7 0.6


Interbirth intervala 7 21.6 — 7 21.4 4.4
Litter size at first observationb 49 2.98 0.19 22 2.5 1.1
Litter size at 1 year 38 2.45 0.25 25 1.3 0.5
a
For Amur tigers, for litters that survived 2 months; for Bengal tigers, for litters that survived >2 weeks.
b
Age at first observation in Nepal was –3 months, and in Russia averaged 4.1 + 1.3 months.

suggest that most reproductive and survival para- populations to persist, survival rates of adult females
meters of Amur and Bengal tigers are similar. Age of must be higher than for these other solitary felids,
first reproduction, interbirth interval, and litter size and must average over 84% (Fig. 13.8). In this regard,
at birth (or slightly thereafter) are all similar in Ben- the only two estimates of survival rates for tigers are
gal tigers in Chitwan National Park, Nepal, and Amur not encouraging: Goodrich et al. (2008) reported
tigers in Sikhote-Alin Zapovednik (these being the adult female survival rates of 0.81  0.10 over a 12-
only two complete datasets that exist for tigers in year period based on mortalities of radio-collared
the wild; Table 13.3). The only significant difference animals in Russia, and Karanth et al. (2006) reported
is a nearly 50% reduction in litter size after 1 year in 0.77  0.05 survival rates for an entire population of
Russia compared to Nepal. This difference is partially tigers in Nagarhole National Park, India, using mark-
due to the fact that 43% of cub mortality in our study recapture analyses of camera-trap data. However, as
was directly or indirectly human-caused (Goodrich noted above, the data from Russia included a period
et al. 2008), and such sources of cub mortality were of intense poaching, in the absence of which, female
probably rare in Nepal when studies were conducted. survival rates were higher than the 84% threshold
But this difference may also be associated with great- suggested by Chapron et al. (2008a) for population
er energy demands on female Amur tigers forced to persistence, and the estimate from India includes
travel over larger home ranges in search of dispersed emigration (i.e. emigration and death could not be
prey and greater vulnerability of cubs to predation differentiated) and combined survival rates of tran-
and other mortality factors when mothers are away. sients and adults of both sexes. Undoubtedly, surviv-
In either case, the difference in cub production ap- al rates of adult resident females are considerably
pears to have marginal impact on population growth higher than the averaged population value in Nagar-
rates in model simulations (Karanth and Stith 1999) hole. Thus, despite dramatic differences in many
and appears to be much less important than adult facets of their biology, key demographic parameters
female survival rates in determining population per- and the constraints on population persistence ap-
sistence (Chapron et al. 2008a). pear to be remarkably similar across tiger subspecies
Two additional key parameters in determining in very different environmental settings.
population persistence in large felid populations ap-
pear to be age of first reproduction and interbirth
interval (Chapron et al. 2008a). These two para-
meters greatly influence potential population
Conservation implications
growth rates, and because these parameters are great- Our studies of the northernmost population of tigers
er in tigers than other large solitary felids such as demonstrate the capacity of tigers to adapt to ex-
leopards (Pa. pardus) or cougars, potential growth treme environmental gradients. Variation in prey,
rates are lower (Fig. 13.8). Consequently, for tiger tiger, and human densities as well as human impacts
McDonald 13-McDonal-chap13 Page Proof page 337 28.8.2009 9:14pm

Amur tiger: a case study of living on the edge 337

2008). In Russia, the most unbiased estimate indi-


1.5 u ga
r cates that 83% of mortalities were human-caused,
Co with poaching the dominant factor (Goodrich et al.
1.3

ard 2008). A similar situation probably prevails across


op
Asymptotic growth rate

Le much of Asia, and direct human-caused mortality is


er likely to be the crucial determinant of the tiger’s
1.1

Tig
future. While relatively high mortality rates of cubs
and subadults can be tolerated, mortality rates ex-
0.9

ceeding 16% for resident, breeding females is likely


to lead to population declines, and ultimately to
0.7

extinction (Chapron et al. 2008a). Age of first repro-


duction and interbirth intervals—variables that ap-
pear to be more or less physiologically and/or
0.5

ecologically constrained—are high in comparison


0.5 0.6 0.7 0.8 0.9 1.0 to other large felids, and consequently restrict the
Breeder survivorship
tiger’s ability to respond to high mortality rates by
Figure 13.8 Asymptotic growth rates (lambda) of increasing birth rates. These findings are paralleled in
cougars, leopards, and tigers as a function of adult breeding lions, the other large felid of approximately the same
female survival rates and species-specific demographic size, which appear to be able to sustain only limited
characteristics (age of first reproduction and interbirth levels of human-caused mortalities (Whitman et al.
interval), derived from a branching process model designed
2004). Tiger populations do not appear to be nearly
specifically for solitary felids (Chapron et al. 2008a).
as ‘resilient’ to human perturbations as has previous-
are all parameters that influence appropriate conserva- ly been thought (Sunquist et al. 1999)
tion strategies for tigers. Here we consider two key Prey recovery efforts are essential for recovery of
components of tiger conservation: first, how human tiger populations across much of the remaining suit-
impacts that directly affect mortality influence viabili- able habitat in Asia (Karanth and Stith 1999; Karanth
ty of tiger populations, and secondly, how to consider et al. 2006), and it is clear that, in the absence of high
land-use planning (habitat conservation) for tigers rates of human-caused mortality, tiger density is di-
given the range of conditions under which they live. rectly related to prey abundance (Figs 13.4 and 13.5a;
These components address what we believe to be the Karanth et al. 2004c). However, because humans are
two key threats to survival of tigers: poaching and the primary mortality agent of tigers, reduction of
habitat loss. Comparing conservation approaches in human-caused mortality is an essential prerequisite
the Indian subcontinent versus north-east Asia is use- to their conservation. Existing tiger conservation
ful in that we consider the two ends of the spectrum of strategies focusing solely on landscape planning (Mi-
human–tiger relationships: one (India) where islands quelle et al. 1999a; Wikramanayake et al. 2004) or
of habitat with tigers are surrounded by a sea of hu- prey recovery (Karanth and Stith 1999) alone may
manity, and the second (Russia) where isolated ‘is- not be sufficient if mortality rates of adult breeding
lands’ of human settlements are surrounded by vast females exceed 16%. Despite evidence of some varia-
forested habitats of tigers. tion in reproductive success between Amur tigers
and other populations, the key parameters that
limit responses by tigers (age of first reproduction
and interbirth intervals) appear similar across their
range (Table 13.3). The importance of reducing
Anthropogenic impacts as mortality
human-caused mortality on adult females appears
agents to be universal for tiger populations across Asia.
Although little information exists, the vast majority Localized extinctions of tiger populations in India
of tiger mortalities appear to be human-caused (Dinerstein et al. 2007) demonstrate that poaching
(Schaller 1967; Miquelle et al. 2005a; Goodrich et al. can eliminate tiger populations, despite high prey
McDonald 13-McDonal-chap13 Page Proof page 338 28.8.2009 9:14pm

338 Biology and Conversation of Wild Felids

densities and the potential for high tiger densities. to provide source populations. Therefore, a success-
Hence, control of poaching and other sources of ful tiger conservation strategy in north-east Asia will
human-caused mortality must be at the forefront of depend upon creating and effectively managing a
tiger conservation initiatives across their range. core network of protected areas that are connected
to and interspersed with multiple-use lands where
tiger conservation is integrated with sustainable use
of natural resources by humans (Miquelle et al.
Habitat loss and fragmentation 1999b, 2005a; Miquelle and Zhang 2005).
Like tigers elsewhere, Amur tigers prey on medium- Although there are more tigers in India than in any
sized and large cervids and wild boar, although wild other country, its burgeoning human population has
cattle, an important component of the diet in some fragmented remaining tiger populations into small,
parts of Asia, are absent in the north. Because plant isolated pockets of habitat. The total amount of land
productivity declines with latitude, carrying capacity available for tiger conservation is quite similar in
for these preferred prey in northern temperate forests north-east Asia (Russia and Northeast China) and
is reduced by an order of magnitude in comparison the Indian subcontinent—271,298 versus 227,569
to southern areas. Prey biomass estimates in high- km2—but in north-east Asia there exist two TCLs
quality habitats in India range from 2000 to nearly (Sikhote-Alin and east Manchurian mountains) that
7500 kg/km2, whereas in the Russian Far East, quality average 135,649 km2, while in the Indian subconti-
habitat supports a prey biomass of less than 600 kg/ nent there are 40 TCLs that average only 2100 km2 in
km2 (Fig. 13.4). Consequently, female Amur tigers size (Sanderson et al. 2006). Of course, the potential
maintain home ranges that are more than 20 times number of tigers that can be held in these landscapes
larger than those of Bengal tigers to find sufficient is dramatically different. Ranganathan et al. (2008)
food to raise young (Fig. 13.6), and their birthing suggest that with effective management both inside
season has shifted to increase hunting success protected areas and in adjacent matrices, over 6000
when demands on females with young are greatest. tigers could be retained on the Indian subcontinent,
Given these constraints, tiger densities in the Rus- but in a more likely scenario where adjacent land
sian Far East rarely exceed one animal/100 km2, matrices are hostile for tigers, only around 3600
whereas some parts of India can boast of more than tigers would survive (provided protected areas are
16 tigers/100 km2. well managed). In north-east Asia, tiger densities
The conservation implications of these ecological could increase if effective ungulate management is
differences are vast. Within the Indian subcontinent, practiced outside protected areas, and poaching is
tiger conservation is largely predicated on retaining eliminated within protected areas. However, densi-
viable populations within individual protected areas ties above 1.0 tigers/100 km2 are likely to occur only
(e.g. Thapar [1999]). Such a tactic would be impossi- in the most productive habitats, and numbers ex-
ble in the north. At 4000 km2, Sikhote-Alin Zapoved- ceeding 1000 Amur tigers are unlikely.
nik is the largest protected area in Russia within tiger At least in the short term, the focus on protected
range, yet harbours fewer than 30 animals, at least areas has been successful in retaining tigers in the
half of which regularly use areas outside its bound- Indian landscape, although this strategy is a risky
aries. A similarly sized reserve in India could contain one, as the probability of extinction for any one of
640 tigers in an ideal setting. Because Amur tigers those isolated populations is relatively high because
require vast home ranges, no single protected area survival is dependent upon strong government regu-
can retain a viable population of tigers. Yet protected lation and compliant people—neither of which is
areas do provide a haven for tigers and their prey. On likely to be continuous. Ranganathan et al. (2008) sug-
unprotected lands adjacent (and presumably similar) gest that survival of tigers in the vast majority of these
to zapovedniks, tigers occur at 50% of the densities landscapes will be dependent on how the environment
observed in zapovedniks because prey densities are surrounding protected areas is managed: under less
lower and anti-poaching efforts are generally weaker favourable conditions in the surrounding areas, the
(Miquelle et al. 2005b). Thus the reserve network acts estimated median population size for 129 of the 150
McDonald 13-McDonal-chap13 Page Proof page 339 28.8.2009 9:14pm

Amur tiger: a case study of living on the edge 339

protected areas is only 8 tigers. As this seems to be the patible with tiger conservation must be initiated
scenario developing in India, in the near future, we are across a vast forested landscape, making the task a
likely to witness many localized extinctions of the kind more complicated one. However, in India, the de-
experienced in Sariska National Park (Dinerstein et al. mands of a swelling population of 1.1 billion people
2007). But even with more effective landscape manage- make the task there more daunting, while in Russia
ment in adjacent territories, which would increase me- the total human population surrounding tiger habi-
dian population size to 14 (Ranganathan et al. 2008), tat is less than 2 million, with the vast majority
the long-term prospects of these small populations is concentrated in a few cities and peripheral agricul-
bleak, as genetic drift and inbreeding depression would tural lands. Hence, the scale of the landscape is vast
likely become major concerns, and active manipula- in Russia, but the scale of problems associated with
tion of populations would be necessary to retain the the human population is much greater in India.
genetic integrity of the Bengal subspecies. Ranga- In Northeast China, a surprising amount of habitat
nathan et al. (2008) suggest that there are 21 protected still remains (Fig. 13.2), presently largely devoid of
area clusters that are sufficiently large to retain viable tigers. While more fragmented than nearby Russia,
populations of tigers even in a hostile environment, with a denser human population exerting pressures
and it is these areas that are probably the last hope for on the landscape, there nonetheless still exists the
tigers in the Indian subcontinent. potential for recovery of tiger numbers in Northeast
In contrast to the Indian subcontinent, tiger habitat China across a broad swath of habitat along the
in the Russian Far East is largely unfragmented, al- provincial boundaries of Jilin and Heilongjiang.
though the risk of future fragmentation exists (Carroll This process is underway with the recovery of tigers
and Miquelle 2006). Nonetheless, despite the fact that in the Hunchun Tiger and Leopard Reserve along the
Amur tigers require home ranges exceeding 400 km2, Sino-Russian border in Jilin Province (Fig. 13.1). Al-
at approximately 430–500 tigers (Miquelle et al. 2007), though it will take time, there exists a great opportu-
Russia retains what appears to be the largest existent nity to increase substantially the total population of
single population of tigers in the world in a single Amur tigers, and develop a viable population, sepa-
unbroken tract of 192,000 km2 of forests, of which rate from the existing Sikhote-Alin population, in
approximately 128,000 km2 (66%) is potential tiger the East Manchurian Mountains (Miquelle et al., in
habitat (Miquelle et al. 1999a). Although future devel- press). Thus, although threats to survival of the Amur
opment and timber harvest is unavoidable (but can be tiger continue to loom, those threats are understood
managed to reduce impact), a declining human popu- and rectifiable, if there is sufficient political will and
lation across Russia and specifically an exodus of peo- constraint of human demands on the landscape.
ple from the forest villages within tiger habitat provide Based on the wealth of legend and lore that has
hope that this landscape may not undergo serious grown up around the tiger in both Russia and North-
fragmentation in the near future. east China, it is clear that not only the natural land-
One similarity that arises from this comparison of scape, but also the human landscape will be the
conservation strategies is the urgent need in both poorer if this population disappears. Yet there is suf-
India and north-east Asia to improve management ficient reason for optimism that this northernmost
outside protected areas to increase the effective representative of Pa. tigris will continue to roam the
amount of habitat for tigers across the greater land- forests of north-east Asia, interacting with natural
scape matrix. In India, the task is on the one hand prey and exposed to the full spectrum of natural
easier because the scale (total amount of land sur- forces that it has survived in for so long, if humans
rounding protected areas) is much smaller; improved can provide the minimum requirements of space,
management on relatively small parcels of land can undisturbed habitats, and freedom from direct
greatly increase total population sizes in isolated human persecution. It is a simple recipe, if we have
TCLs. In Russia, multiple-use scenarios that are com- the collective societal will to make it happen.
McDonald 13-McDonal-chap13 Page Proof page 340 28.8.2009 9:14pm

View publication stats

You might also like